JPH11262010A - Error permissible mode video codec method and is device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an error permissible mode video codec method, in which the information loss is small and a large error permissible range is added. SOLUTION: This method includes a step 20, where each macro block area of video data in the error permissible mode are divided into a header data part bit area, a motion vector data part bit area and a discrete cosine transform data part bit area, a step 22 where variable length coding is applied to the divided bit areas, a step 24 where inverse variable length coding is applied to the bit area selected based on a prescribed priority which is required for decoding among the bit areas variable length coded, and a step 26 where a marker is inserted to the bit areas that are variable-length coded or inverse variable-length coded.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a video codec (Co-codec).
More particularly, the present invention relates to an error-tolerant mode video codec method and apparatus.
The present invention also provides a video codec method in an "error tolerant mode" filed by the assignee of the present application.
Method in Error Resilient mode), based on US Provisional Patent Application No. 60 / 067,013.

[0002]

2. Description of the Related Art Generally, in a situation where a channel has an error, a video codec has a reduced ability to cope with a channel error in encoding an entire frame into a bit stream. This is because there is no decision factor to determine whether the resulting portion should be restored or discarded. Therefore, it is difficult to determine whether a certain part of a frame has a bit loss due to an error or an error in the whole, so that the coded bit stream of the entire frame is ignored and the bit for the next frame is ignored. The drawback was that the starting point of the stream had to be found.

In order to solve this problem, a codec according to the prior art uses an additional device separately and uses it as a judgment factor. As an example, H.S. In the codec defined in the H.263 standard, a picture start code (PSC: Picture Start Code) is placed, and if an error is received next to this field information, the subsequent part is ignored and the next picture start code (PSC) is ignored. The method of searching for is used. A block group (GOB: Group Of Bloc)
Block start code (GBS) indicating the start of k)
C: GOB Start Code), and if an error is received next to this information, only the subsequent part is ignored and the next block group is searched to reduce the discarded part.

FIG. 1 shows an example of a video data packet generated in an error-tolerant mode codec according to the prior art. The video data packet shown in FIG. 1 is an example of a video data packet generated in the MPEG-4 codec. Referring to this packet, the error-tolerant mode codec according to the related art has motion data and texture data. Encoding is performed separately from data (texture data). The motion data includes a COD bit indicating whether coding is possible and a macroblock pattern chromaticity (MCBPC: MaCro Block Pattern) indicating the chromaticity type of each macroblock.
Chrominance) bits and motion vectors, texture data is CBPY (Coded Block Patt)
ern luminance (Y)), DQUANT (Data Quantizatio
n) and DCT (Discrete Cosine Transform) data. The above-mentioned "COD" is a field indicating the presence or absence of encoding in the bit stream, and is an abbreviation indicating a code. The motion data and the texture data are motion markers (MM).
Is separated into In addition, DCT data is a reversible variable length coding (RVL).
C).

Next, a process of decoding the coded bit stream will be described.
First, a resynchronization marker (RM) is found in the bitstream. Once the resynchronization marker is found, the information until the next resynchronization marker is found is regarded as one packet. Motion vector data part (MVDP: Motion Vector Data)
Since only the motion vector by prediction exists in Part), the motion vector can be actually used only when there is a previously decoded motion vector. Therefore, if an error is found in the motion vector data portion in the bit stream, the corresponding entire packet is ignored, and a resynchronization marker corresponding to the next packet should be found and decoded. There is a disadvantage that there are many.

[0006]

SUMMARY OF THE INVENTION It is an object of the present invention to provide an error tolerant mode video codec method which can reduce the loss of information and greatly add error tolerance. Another object of the present invention is to provide a computer-readable recording medium for executing the video codec method. It is still another object of the present invention to provide an error tolerant mode video codec device that implements the method.

[0007]

In order to achieve the above object, a video codec method according to the present invention comprises a header data part bit area, a motion vector data part bit, and a macro data area from each macro block area of the video data in an error tolerant mode. Dividing the divided bit area into variable-length coded data areas, and a step of performing variable-length coding on the divided bit areas, based on a predetermined priority required for restoration of the variable-length coded bit areas. And performing a variable length coding on the selected bit region, and inserting a marker into the variable length or inverse variable length coded bit region.

According to another aspect of the present invention, there is provided a video codec apparatus comprising: a header data portion bit region, a motion vector data portion bit region; Means for dividing the discrete cosine transformed data portion into bit areas; means for performing variable length coding on the divided bit areas; and selecting the variable length coded bit areas based on a predetermined priority necessary for restoration. And a means for inserting a marker into the bit area subjected to the variable length coding or the inverse variable length coding.

DETAILED DESCRIPTION OF THE INVENTION Hereinafter, a preferred embodiment of the present invention will be described in detail with reference to the accompanying drawings. FIG. 2 shows the main steps of an error-tolerant mode video codec method according to an embodiment of the present invention. Referring to FIG. 2, the video codec method in an error-tolerant mode according to the present invention includes a dividing step 20, a variable-length coding step 2,
2, an inverse variable length coding step 24 for the selected bit region and a marker insertion step 26.

In the dividing step 20, a header data part (HDP:
Header Data Part) bit area is formed, and similarly, a motion vector data portion (MVDP:
A Motion Vector Data Part (DCP) bit area and a DCT data part (DDP: DCT Data Part) bit area are configured.

[0011] The header data contains information about the coded state of the current macroblock. That is, the header data indicates whether the current macroblock is an inter macroblock obtained by coding the contents of the current frame as it is,
Or, it indicates whether it is an intra macroblock in which the difference of the previous frame is coded, and if an error occurs in the header data, extremely serious information loss is caused. Therefore, this header data has a high importance and should be restored preferentially when restoring. Thereby, as an example, the macroblock identification (COD) bit and the macroblock pattern chromaticity (MCBPC) bit indicating the coding availability of the header data are combined into one new syntex (COD + MCBPC) bit. It is desirable to divide it. The reason why it is preferable to combine and divide this way is as follows.

The present H. In the H.263 standard, 1 bit is used for the COD bit and a variable length code is used for the MCBPC bit as header data. However, since the COD bit can be either 0 or 1, it is difficult to identify whether an error has occurred. In addition, even in the case of MCBPC bits using a variable length code, if an error occurs in a received bit, it is highly possible that the error bit exists in the variable length code table, so that error detection is difficult. . Therefore, according to an embodiment of the error tolerant mode video codec method according to the present invention, the COD bit and the MCBPC bit are combined at the time of division, and the combined bit region is inverse-variable-length coded,
It is easier to identify whether a bit has changed.

For example, in the case of 0110, even if an error such as 0101 or 1001 occurs, it can be identified as an error, so that the possibility of using information containing the error can be reduced. Moreover, since the coding is performed by the inverse variable length coding, the restoration in the reverse direction is also possible. However, if necessary, the header data portion may be constituted by a single division of the header data without using the combining method.

Further, since the dividing step 20 is performed for each macroblock, a step of inserting a first macroblock index bit region indicating the number of the first macroblock into each packet is included.
More preferably, it can be referred to at the time of decoding. Preferably, the dividing step 20 includes a step of inserting a final absolute motion vector bit including absolute motion vector information, which is not based on prediction, after the motion vector data part bit area. Further, the dividing step 20 preferably further includes a step of inserting a packet number bit area indicating a serial number of the packet as additional information.

In the variable length coding step 22, a variable length coding operation is performed on the bit region of the header data portion, the bit region of the motion vector data portion, and the bit region of the DCT data portion. In addition, the packet number and the final absolute motion vector bit are also variable-length coded.

In the inverse variable length coding step 24, inverse variable length coding conversion is performed on a bit area selected based on a predetermined priority necessary for restoration among the variable length coded bit areas. That is, at the time of decoding, inverse variable length coding is performed on important information that can cause loss of the entire packet or macroblock if a loss occurs. The reason why the priority order is necessary will be described in detail below.

First, as described above, the header data indicates whether the current macroblock is an inter macroblock in which the contents of the current frame are coded as it is or an intra macroblock in which the difference between the previous frame is coded. Most macroblocks are composed of inter macroblocks to increase compression efficiency. If an error occurs in the header section in the inter macro block, the motion vector section and DCT data section cannot be used, and if an error occurs in the motion vector section, the DCT data section cannot be used. Therefore, importance is assigned in the order of the header section, the motion data section, and the DCT data section. That is, there is a priority corresponding to such importance in the macroblock, and it is necessary to restore the macroblock according to the priority. At this time, it is not necessary to perform inverse variable length coding on all data, but to perform reverse variable length coding selectively with priorities. Is increased, and a limited channel capacity is considered.

Although not shown in FIG. 2, the video codec method according to the present invention uses the channel capacity, the error situation,
The method further includes a step of identifying channel characteristics by receiving information on channel characteristics such as congestion through a back channel or the like. If the channel characteristics allow in the channel characteristic identification step, for example, The method further includes a step of performing inverse variable length coding on a predetermined bit area having a low priority such as a bit area of a DCT data part and performing variable length coding on the bit area when channel characteristics are not permitted, thereby increasing error tolerance. However, it is more desirable to realize the characteristics of the channels while satisfying them.

Further, as an alternative, if the channel characteristics allow in the channel characteristics identification step, for example, D
Additional information is inserted into a bit area having a low priority, such as a bit area of the CT data section, so that it can be referred to at the time of decoding.

Hereinafter, the use of such additional information will be described in detail with an example. First, the information of the header part is
It is possible to send it again into the data section. This is for the case where an error occurs in the header section with high importance and no error occurs in the motion vector section and the DCT data section, and this situation may actually occur. By locating the header data after the DCT data portion in this way, it is possible to restore the current data packet without discarding it. Second, it is possible to send a CRC as additional information in the DCT data area. For example, DCT
The data is an 8-bit fixed-length code (FL) as a DC value.
C) uses a variable length code (VLC) as an AC value. An 8-bit fixed-length code is relatively difficult to detect an error. By adding CRC to this 8-bit fixed-length code to assist in error detection, the resilience can be improved. Further, other bits using a fixed length code together with the 8-bit fixed length code,
For example, QP bits indicating a quantized value can be used in creating a CRC. In this way, DC
In the T data portion, CRC may be added to the fixed length code bits to improve error recovery at the time of decoding.

After all, the additional information in the above two cases is
This is to make it possible to recover a portion where an error has occurred in the current packet or to make it easier to detect a portion where error detection is difficult. In addition, in the error tolerant mode video codec method according to an embodiment of the present invention, when inverse variable length coding is performed, the RVL having different characteristics depending on channel characteristics in the channel characteristic identification step.
Error detection is performed more efficiently by using the C coding table.

In the marker insertion step 26, a resynchronization marker for dividing the packet and a header marker as information bits for separating the header and the motion vector are inserted. Further, in the marker insertion step 26, a motion marker for distinguishing between the motion vector data part and the DCT data part is inserted. Since the header marker uses a code word that is not used for coding the header data portion, it is desirable that the header marker can be distinguished at the time of decoding.

FIG. 3 shows an error-tolerant mode video codec according to an embodiment of the present invention for implementing the above method. Referring to FIG. 3, the error tolerant mode video codec according to the present invention is configured based on a dividing unit 30, a variable length coding unit 32, and a predetermined priority required for restoration among the variable length coded bit regions. Inverse variable length coding means for inverse variable length coding of the selected bit region is provided. In addition, the error tolerant mode video codec device identifies the channel characteristics by receiving information on the channel characteristics such as channel capacity, error status, and congestion level, for example, via a back channel (not shown). And a marker insertion unit 38 for inserting a marker into a bit region that has been subjected to variable length coding or inverse variable length coding.

The dividing means 30 inputs video data,
As described with reference to FIG. 2, in the error tolerant mode, the header data part (HDP) bit area, the motion vector data part (MVDP) bit area, and the discrete cosine transform data part (DDP) ) Divide into bit areas. Further, the dividing means may convert the header data portion bits into, for example,
A plurality of predetermined bit areas having high importance in restoration, such as a coding state (COD) bit and a macroblock pattern chromaticity (MCBPC) bit, are combined to form one new syntex (COD + MCBPC) bit. desirable.

The variable-length coding means 32 performs variable-length coding by inputting the divided bit areas, and the inverse variable-length coding means 34, as an example of the variable-length coded bit areas, which is lost when decoding. If this is the case, inverse variable length coding conversion with good error recovery is performed on a bit region selected based on a predetermined priority required for restoration, such as the possibility that the entire packet or macroblock may be lost. However, if the inverse variable length coding is performed, the number of bits is increased as compared with the case where the variable length coding is performed. Therefore, the priority is set in consideration of the limited channel capacity, and the selected bit area is assigned a priority. It is desirable to perform the inverse variable length coding only when the coding is performed.

Channel characteristic identification means 36 for that purpose
If the channel characteristics are determined to be acceptable, inverse variable length coding is performed on a bit region having a low priority, such as a discrete cosine transform data portion, and if the channel characteristics are determined not to be permitted, the bit region is rewritten. The variable length coding means 32 and the inverse variable length coding means 34 are controlled so as to perform variable length coding. Alternatively,
A means (not shown) for inserting additional information is further provided, and when it is determined by the channel characteristic identification means 36 that the channel characteristic is permitted, the additional information is added to a bit area having a low priority such as a DCT data part. It is also possible to insert. Further, the inverse variable length coding means 3
When the channel characteristic is determined to be acceptable by the channel characteristic identification means, it is preferable to perform inverse variable length coding using another coding table in order to achieve the effect described based on FIG. .

Further, the video codec according to the present invention, in order to achieve the effect described with reference to FIG. 2, uses a first macroblock index (FM) indicating the number of the first macroblock in each packet.
A means for inserting a BI (First Macro Block Index) bit area (not shown) is further provided, and the inverse variable length coding means 34 includes a final absolute motion vector (LAMV:
(Last Absolute Motion Vector) bits are desirably subjected to inverse variable length coding. Further, the video codec device may further include a packet number (PN: Packet Number) indicating a serial number of the packet next to the resynchronization marker bit area.
r) means for inserting a bit area as additional information (not shown), and means for inserting a final absolute motion vector bit including absolute motion vector information next to the motion vector data part bit (not shown). It is desirable to include.

On the other hand, the marker insertion means 38 includes a resynchronization marker (RM) for separating the packet and a code word not used for coding the header data part as information bits for separating the header part and the motion vector part. A header marker (HT) used and a motion marker (MM) for separating the motion vector data portion from the DCT data portion are inserted. At this time,
By using a code word not used for coding the header data portion, the header marker,
It is desirable to be able to distinguish when decoding.

FIG. 4 shows an example of a video data packet generated by the above method. Referring to FIG. 4, a video data packet generated by the above method includes a resynchronization marker (RM) bit and a packet number (P).
N) bits and the first macroblock index (F
MBI) bit. The packet includes a header data part (HDP) bit, a header marker (HM) bit, and a motion vector data part (MVDP) bit. The packet includes a final absolute motion vector (LAMV) bit, a motion marker (M
M), and a DCT data part (DDP).

In the video data packet thus configured, the resynchronization marker bit indicates the start of the video data packet, so that the decoding start point can be found during decoding, as in the prior art. . The packet number indicates the serial number of each packet and can be used as additional information. The first macroblock index bit indicates the number of the first macroblock in each packet. In the present embodiment, the header data part bit is MPEG-4 or H.264. The coding state (COD) bits defined in the H.263 standard and the macroblock pattern chromaticity (MCBPC) bits are combined as one new syntex (COD + MCBPC) bit, and the bits are inverse-variable-length coded. The header marker bits are information bits for distinguishing between the header part and the motion vector part, and include a code word that is not used for COD + MCBPC coding of the header data part bits.

The motion vector data portion bit is a motion vector data portion obtained by coding information generated by the motion vector prediction by an inverse variable length coding method. The DCT data part bit is an area including information related to the discrete cosine transform, and CBPY
Bits, DQUANT bits, and DCT coefficients are coded. The DCT data part bits include:
It is desirable to insert information necessary for error confirmation and restoration in decoding after reception among information used in the current packet. The insertion of such information can also be selectively applied depending on the characteristics of the channel. The motion marker bit is the motion vector data part bit and the DCT
This is for distinguishing from the data part bit.

Hereinafter, a process of decoding the video data packet encoded as described above will be described. First, when a resynchronization marker is found in the received bit stream, it is combined into one packet until the next resynchronization marker is found, as in the prior art. If an error occurs in the header data part, the header data part is coded by the inverse variable length coding method, so that not only decoding in the forward direction but also decoding in the reverse direction becomes possible. For example, the header data part can be restored using the packet number of the next packet and the first macroblock index. That is, since the value obtained by subtracting 1 from the first macroblock index of the next packet is the macroblock number of the current packet, it is possible to restore the header data portion bits to the area where the error has occurred by using the backward decoding. Further, the motion vector data part bits are decoded until a motion marker is found, thereby decoding a motion vector based on the motion vector prediction. If an error occurs in the decoding process, the motion vector data portion bits are also inversely variable-length coded, so that decoding in the reverse direction becomes possible.

This will be described in comparison with a process of decoding a video data packet encoded by the conventional codec method. In the conventional method, only the motion vector obtained by the motion prediction is included in the motion vector data part bit. As such, the substantial motion vectors were only available if there were previously decoded motion vectors. However, in the codec method according to the present invention, an absolute motion vector that does not predict the final motion vector is coded and transmitted by the inverse variable length coding method. When decoding, it is independent of the previous motion vector.
Further, if an error is received in the DDP bit area, the DDP bit area has to ignore the packet if it is coded by variable length coding, but it has to reverse the data if it is coded by inverse variable length coding. Since coding can be performed, there is an effect that error tolerance is further given. If the DDP bit area encodes additional information necessary for restoration, it can be used for decoding.

On the other hand, the above-described embodiment of the present invention can be created by a program executable on a computer. Further, the present invention can be embodied by a general-purpose digital computer that operates the program from a medium used in a computer. The medium includes a storage medium such as a magnetic recording medium (eg, floppy disk, hard disk, etc.), an optical recording medium (eg, CD-ROM, DVD, etc.), and a carrier wave (eg, transmitted via the Internet).

The recording medium divides each macroblock area of the video data into a header data part bit area, a motion vector data part bit area, and a discrete cosine transform data part bit area in an error tolerant mode; Performing variable length coding on the selected bit region, performing inverse variable length coding on a bit region selected based on a predetermined priority required for restoration among the variable length coded bit regions, and variable length coding Alternatively, a program code for performing a video codec method including inserting a marker into a bit area coded by inverse variable length is stored.

[0036]

As described above, the codec method according to the present invention can further provide an error-tolerant mode, so that communication that is less sensitive to errors in an environment where errors are greatly generated such as a wireless communication channel. Will be possible. As described above, the codec method according to the present invention uses MPEG
-4 and H.E. Although H.263 has been described as an example, the terms related to the bit area such as the resynchronization marker, the packet number, the first macroblock index, the header data part, the header marker, the motion vector data part, the motion marker, and the DCT data part are defined and described. As will be apparent to those skilled in the art, it is applicable to any video codec that uses other bit regions equivalent to those defined in this term. Further, functional programs, codes, and code segments for embodying the present invention can be easily inferred by programmers in the technical field to which the present invention belongs.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating an example of a video data packet generated in an error-tolerant mode codec according to the related art.

FIG. 2 is a flowchart illustrating main steps of an error tolerant mode video codec method according to an embodiment of the present invention.

FIG. 3 is a block diagram illustrating an error tolerant mode video codec device according to an embodiment of the present invention.

FIG. 4 is a diagram illustrating an example of a video data packet generated by an error tolerant mode video codec method according to an embodiment of the present invention.

[Explanation of symbols]

20 segmentation step 22 variable length coding step 24 inverse variable length coding step for selected bit region 26 marker insertion step

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor ▲ Wen ▼ Jato United States, 90095-1594, California Los Angeles, Los Angeles University of California, Hilgard Avenue 405, Electrical Engineering・ Department (72) Inventor John Di Wilsner United States 90095-1594 ・ California Los Angeles Los Angeles University of California Hilgard Avenue 405 ・ Electrical Engineering Department (72) Inventor Park Tae Hun

Claims (29)

[Claims] 1. A video codec method for coding video data to generate a video data packet and decoding the packet, comprising: a header data part bit area, a motion vector and a header data part from each macroblock area of the video data in an error tolerant mode. Dividing the bit region into a data portion bit region and a discrete cosine transform data portion bit region; performing a variable length coding on the divided bit region; and a predetermined priority required for restoration of the variable length coded bit region. A video codec method, comprising: performing inverse variable length coding on a bit region selected based on an order; and inserting a marker into the variable length or inverse variable length coded bit region. 2. The video codec method according to claim 1, wherein the inverse variable length coding step performs an inverse variable length coding on a bit area of a header data part and a bit area of a motion vector data part. 3. The video codec method according to claim 1, wherein the header data portion bits form one information bit by combining a plurality of predetermined bit regions having high importance in restoration. 4. The predetermined bit area includes a macroblock identification (COD) bit and a macroblock pattern chromaticity (MCBPC) bit, which are important for restoration, and the COD bit and the MCBPC bit are combined to form one bit. 4. The video codec method according to claim 3, wherein a new syntex (COD + MCBPC) is formed. 5. The video codec method according to claim 1, wherein the dividing step includes a step of inserting a first macroblock index bit region indicating a number of a first macroblock in each packet. . 6. The video codec according to claim 2, wherein the dividing includes inserting a final absolute motion vector bit including absolute motion vector information after the motion vector data part bit. Method. 7. The video codec method according to claim 6, wherein the final absolute motion vector bits are inversely coded. 8. The video codec method according to claim 1, further comprising a step of inserting, as additional information, a packet number bit area indicating a serial number of the packet after the resynchronization marker bit area. 9. The step of inserting a marker is a step of inserting a resynchronization marker for identifying a packet, and is not used for coding a header data part as information bits for identifying a header part and a motion vector part. The method of claim 1, further comprising: inserting a header marker using a codeword; and inserting a motion marker for distinguishing between a motion vector data part and a discrete cosine transform data part. Video codec method. 10. The method of claim 7, further comprising the step of identifying a characteristic of the channel, wherein the step of identifying the characteristic of the channel is performed by inverse variable length coding of a predetermined bit area having a low priority if the characteristic of the channel permits, and the characteristic of the channel does not permit. 2. The video codec method according to claim 1, further comprising a step of performing variable length coding on the bit area. 11. The video codec method according to claim 10, wherein the bit area is a discrete cosine transform data part. 12. The method of claim 11, further comprising: identifying a channel characteristic; and inserting additional information into a predetermined bit area having a lower priority if the channel characteristic permits in the channel characteristic identification step. The video codec method according to claim 1, wherein 13. The video codec method according to claim 12, wherein the bit area is a DCT data part. 14. The method of claim 11, further comprising the step of identifying characteristics of the channel, wherein, if the channel characteristics allow in the identifying of the channel characteristics, using different coding tables when performing the inverse variable length coding in the encoding. The video codec method according to claim 1, wherein: 15. A computer readable recording medium that executes a video codec method of coding video data to generate video data packets and decoding the packets, wherein each macroblock of the video data is in an error tolerant mode. Dividing the bit region into a header data portion bit region, a motion vector data portion bit region, and a discrete cosine transform data portion bit region, and performing variable length coding on the divided bit region; and Performing a step of performing inverse variable length coding on a bit area selected based on a predetermined priority necessary for restoration among the areas and a step of inserting a marker into the variable length coded or inverse variable length coded bit area; program Recording medium that can be read by a computer that you want to save. 16. A video codec apparatus for generating video data packets by coding video data and decoding the packets, comprising: a header data part bit area, a motion vector, Means for dividing into a data part bit area and a discrete cosine transform data part bit area; means for performing variable length coding on the divided bit area; and a predetermined priority required for restoration of the variable length coded bit area. A video codec device comprising: means for performing inverse variable length coding on a bit area selected based on the order; and means for inserting a marker into the variable length coding or inverse variable length coded bit area. 17. The video codec according to claim 16, wherein the inverse variable length coding means performs an inverse variable length coding on a bit area of a header data part and a bit area of a motion vector data part. 18. The apparatus according to claim 16, wherein said dividing means combines a plurality of predetermined bit areas of high importance to form one information bit when restoring said header data part bit. Video codec equipment. 19. The predetermined bit area includes a macro block identification (COD) bit and a macro block pattern chromaticity (MCBPC) bit, which are important for restoration, and the COD bit and the MCBPC bit are combined to form one bit. New Syntex (COD + MCBPC)
19. The video codec device according to claim 18, wherein 20. The video codec apparatus according to claim 16, further comprising means for inserting a first macroblock index bit area indicating the number of the first macroblock in each packet. 21. The video codec according to claim 16, further comprising means for inserting a final absolute motion vector bit including absolute motion vector information after the motion vector data part bit. 22. The video codec according to claim 21, wherein the inverse variable length coding means performs an inverse variable length coding on the final absolute motion vector bit. 23. The video codec according to claim 16, further comprising means for inserting, as additional information, a packet number bit area indicating a serial number of the packet after the resynchronization marker bit area. 24. The marker insertion means uses a resynchronization marker for dividing a packet and a codeword not used for coding of a header data part as information bits for dividing a header part and a motion vector part. The video codec device according to claim 16, wherein a header marker and a motion marker for distinguishing between a motion vector data portion and a discrete cosine transform data portion are inserted. 25. The apparatus according to claim 25, further comprising means for identifying a characteristic of the channel, wherein the channel characteristic identifying means performs inverse variable length coding on a predetermined bit area having a low priority when the channel characteristic is determined to be acceptable, and determines whether the channel characteristic is acceptable. 2. The method according to claim 1, wherein when it is determined not to perform the coding, the bit area is controlled to perform variable length coding.
7. The video codec device according to 6. 26. The video codec according to claim 25, wherein the bit area is a discrete cosine transform data section. 27. A system further comprising: means for identifying a characteristic of a channel; and means for inserting additional information into a predetermined bit area having a low priority when the channel characteristic is determined to be acceptable by the channel characteristic identifying means. 17. The video codec device according to claim 16, wherein: 28. The video codec according to claim 27, wherein the bit area is a discrete cosine transform data section. 29. An apparatus according to claim 29, further comprising means for identifying a characteristic of the channel, wherein said inverse variable length coding means uses a different coding table if said channel characteristic identification means determines that the channel characteristic is acceptable. 17. The video codec device according to claim 16, wherein: